مجله مهندسی مکانیک امیرکبیر
سال پنجاه و پنجم شماره 2 (اردیبهشت 1402)

Here the healing process of glass/epoxy composites is studied experimentally. The healing agent is composed of ML506 epoxy resin and HA-11 hardner which is charged into microtubes. The microtubes are interleaved between the first and the second and the fifth and the sixth layers. The initial damage is introduced to the samples by drop weight impact tests and the recovery percentage of the tensile strength due to healing process is measured by tensile test. The effects of healing time (without thermal cycles), number of healing units and thermal cycles are investigated on the recovery of the tensile strength. Composite samples containing 8, 16 and 32 healing units are studied in the periods of 1, 6 and 12 days after the initial damage. Also, some damaged samples are set to 1, 3, 5 and 7 thermal cycles and after one day, tensile tests are carried out. The results show that without thermal cycles the healing process is almost completed after 6 days with an 83% recovery. In addition, the maximum amount of tensile strength recovery is equal to 86% which is related to the samples with 32 healing units after 12 days. This amount of healing efficiency can also be achieved by means of 5 thermal cycles. This is also true for thermal cycles where the effect of thermal cycles is tangible up to 5 cycles.

Auxetic materials with negative Poisson’s ratio, as a group of metamaterials, attracted significant attentions among researchers due to their interesting and remarkable mechanical properties. Rigid rotating structures are a subcategory of auxetic materials which can show the same behavior in various directions by tuning their parameters, but due to using rigid rotating blocks their relative density is high. As the rotating blocks are connected by weak joints, stiffness and strength of these structures are low and considering high relative density of these structures specific mechanical properties are even in worse condition. In this research, novel lattice structures based on rigid rotating structures but with remarkably lower relative density were presented. To reduce relative density of these structures, bar elements were used instead of rigid blocks. 3D printing method was used to manufacture samples with these structures and then tensile test was performed on the samples. Poisson’s ratios of the samples were measured by recording image of the structures before and during deformation. The behavior of the structures was predicted by finite element method and compared with experimental measurements. Both of the methods showed auxetic behavior of the structures. Then deformation mechanism of the structures and the effect of the structures shape on the auxeticity were investigated.

According to the wide presence of beams in engineering structures, it is very useful to understand how the beams vibrate nonlinearly in conditions where they oscillate with a large amplitude. In this paper, the nonlinear vibrations of an Euler-Bernoulli beam under finite strain are investigated. In this study, unlike other papers, in order to obtain the governing equations of beam, the field-displacement relationship has been done without approximation. Based on this, the strain-displacement relations are calculated using the Green Lagrange strain and the nonlinear form of the equations is obtained by using the Hamilton method. In order to solve the partial differential equation, using the Galerkin method, the equation has been converted to an Ordinary differential equation and finally solved using the multiple scale method and compared with the Rung-Kutta numerical method. To evaluate the accuracy of the method and the validity of the modeling, the obtained results are compared with the Euler–Bernoulli beam theory and the Von-Karman nonlinear model. The results show that the present method in low vibration amplitudes is consistent with the model of Euler-Bernoulli and Von-Karman, but with increasing amplitude of oscillations, the results of these models will be significantly different from each other, which is as expected.

In this paper, an analytical solution for bending analysis of functionally-graded piezoelectric plate under uniformly-distributed transverse loading with two simply-supported parallel edges and two other arbitrary boundary conditions is presented. The five-variable refined plate theory is employed for describing the displacement. This theory, despite the few numbers of unknown variables, predicts a parabolic distribution for transverse shear stresses across the thickness and also consider the thickness stretching effect. The governing equations are obtained by Hamilton’s principle and Maxwell's equation and these coupled equations is solved using the Levy-type solution in conjunction with the state-space approach. The obtained results are compared with the higher-order shear theories and finite element simulation obtained from Abaqus software which confirms the accuracy and efficiency of the proposed method. It can be seen that for the length-to-thickness ratio of 10 and the power index of 0.5, the value of non-dimensional deflection of the plate with the clamped boundary condition is 0.3327, which has the largest amount of stiffness and the least value of deflection, while the value of the non-dimensional deflection of the plate with two parallel free boundary condition edges is 2.2036 and, as a result, has the lowest amount of stiffness and the highest value of deflection.

In this study, the effect of laminate type and number of layers, fiber angle and modulus of elasticity in combination with the effect of global and local geometric defects has been investigated as a new combination in the field of aeroelasticity. Using the principle of virtual work, by directly integrating the problem-solving boundary, the governing equations are determined based on Kirchhoff thin-plate theory. Then, using the assumption mode method in Galerkin's theory, the partial differential equations are converted to ordinary nonlinear differential equations. The final nonlinear equations are solved using the Runge–Kutta numerical method and the time domain results are extracted to determine the flutter and post-flutter behavior of the plate. The results of the analysis showed that the geometric defect with non-uniform and asymmetric load production, the type of layering, the number of layers and mechanical loads are effective on the plane flutter boundary. The effect of local geometric defects in determining the flutter border is not necessarily destabilizing, but in some cases, depending on the size and location of the defect, it is also possible to increase the stability of the plane flutter boundary. In addition, the dynamic behavior of the plate under the effect of local geometric defects with different shapes and dimensions is very diverse and different from the of general defect (first mode shape) or shell with small curvature.

Parallel manipulators are of interest in various industries due to their high precision, rigidity, high speed and low inertia. Controlling these types of systems faces challenges due to their complex and non-linear dynamics. Among the many methods of controlling the path of parallel manipulators, computed torque and sliding mode methods are the famous methods that are proposed. In practical applications, when the speed of the robot increases, adjusting the controller parameters is very difficult and depends on the working conditions of the robot, so the robot cannot work properly with fixed and predetermined coefficients under any condition. The type of path, the speed of the robot along the path, the initial conditions of the end effector of the robot in relation to the path, and even the sampling time are factors that affect the accuracy of the controller, and by changing each of them, it may be necessary to redefine the parameters of the control system and change the control coefficients. In this article, a method is presented which is based on the sliding mode method and the coefficients of the control system are adjusted appropriately by changing the sliding surface and sliding speed using the fuzzy method. The performance of this method has been investigated in two ways: modeling in MATLAB software and real time applying it to a planar parallel robot.